Spatial clustering of rabies virus genomes using affinity propagation clustering
Sprache des Titels:
Rabies is one of the oldest known zoonosis caused by rabies virus, which is an important species of the genus Lyssavirus. So far, the spread of rabies virus is analyzed on regional levels since a global phylogenetic clustering and classification system is not yet available. Phylogenetic trees of rabies genome sequences calculated by the Maximum Likelihood method suggest a space-dependent clustering. However, these analyses revealed two limitations: (i) The analysis of large datasets results in highly complex dendrograms. (ii) The clustering of phylogenetic trees by visual inspection leads to different results since criteria for cluster definition are still lacking.My presentation aims at showing how these limitations can be solved by means of affinity propagation clustering. This is a mathematical method that is able to uses the phylogenetic distance matrix to allocate sequences to generic clusters. I will present you how affinity propagation clustering was applied to the distance matrices derived from the RABV full genome sample sets, resulting in a cluster structure which strongly corresponds to the structure of the Maximum Likelihood-based phylogenetic tree. At the end of my presentation I would like to discuss on strategies to implement a workflow based on this method to validate evidence for space-dependent clustering of rabies virus sequences.